High-throughput gait acquisition system for freely moving mice.

Area of Science:

• Neuroscience
• Animal Behavior
• Biomechanical Engineering

Background:

• Gait analysis on linear walkways is crucial for distinguishing normal and pathological locomotion in animals.
• Traditional methods are labor-intensive, introduce experimental bias through animal handling, and are time-consuming.

Purpose of the Study:

• To develop and validate an unsupervised, automated system for collecting and analyzing mouse gait metrics.
• To reduce animal handling stress and experimental bias associated with traditional gait analysis methods.

Main Methods:

• A novel system with an embedded runway in an arena allowed unsupervised, ad libitum mouse locomotion.
• DeepLabCut was used for multi-body part tracking, with data processed by the GaitGrapher pipeline for gait metric extraction.

Main Results:

• The unsupervised approach yielded gait parameters comparable to a validated supervised method (Visual Gait Lab).
• The system successfully recorded kinematic data during freely moving mouse locomotion.

Conclusions:

• The automated, unsupervised system is a viable and efficient alternative for collecting kinematic data for gait analysis.
• This approach minimizes animal handling, reduces stress, saves time, and provides reliable gait metrics.